how exactly does @DyalogAPL work? the catbot, that is. seems rather sophisticated for a chatbot!

*chatbot sorry

although robotic cats would be kinda cool

@matt Adám made a secure kernel for running apl stuff online, and linked it to an SE bot.

it might actually be running from tio, not sure

<moon-child> @matt the chatbot bridges two different chatrooms; we're real people, here!

<moon-child> @Razetime the apl stuff is running on the same machine as the bot itself now, so it can return results essentially instantly. (The delay is because of SE chat.)

@Moonchild You're very convincing as a human.

@eyepatch Much obliged

lmoaooo

¯1⍭56

⋄¯1⍭56

@Razetime Illegal code

tio.run/##SyzI0U2pTMzJT9dNrShJzUtJTfn/v/rQesNHvWsf9W6t/Z/… why isn't this returning a list?

@Razetime Is ⍭ function only available in certain Dyalog version?

<moon-child> @jimfan it's part of github.com/abrudz/dyalog-apl-extended

@DyalogAPL So it is said "Requires Dyalog APL version 16.0 or higher."

<moon-child> @jimfan yes, but it's an extension to dyalog. It doesn't come with apl; you have to install it separately

@DyalogAPL Alright

Thanks, bot

<moon-child> ;<

@DyalogAPL I missed this. You said you are real :-)

@jimfan It's in Extended, which has some nifty symbol functions for easy use

best way to use it is tio, I prefer having the Unicode IDE

Heyyy

⋄{(⌽,⊢)⍉↑{⍵↑'O'}¨1,1-1↓+\⍳⍵+1}3

@Razetime
   OO
  O  O


 O    O



O      O

what are some good ways to golf this?

codegolf.stackexchange.com/questions/58732/… also how do i answer questions like these?

<kritixilithos> {⍵↑'O'} → ↑∘'O'

Interesting.. will note that down.

Okay here is my latest golf: codegolf.stackexchange.com/questions/77270/…

@Razetime Much simpler one compare to @Razetime 's

@Razetime Ok thanks for the pointer

nice

@jimfan was trying this in extended, but didnt work

@Adám what's wrong in this one: tio.run/##AVEArv9hcGwtZHlhbG9nLWV4dGVuZGVk//…

@Razetime I think it is the problem with ∩/

It seems to be it evals to not flat vector 2 2 but ⊂2 2

This happens with NARS2000 too

Oops sorry for the ugly output

And not surprising, ×/⊂2 2 gives ⊂2 2

For your case I believe ⍭⍵ gives two boxed array, each being prime factors of 28 and 144, ∩/ reduces them giving ⊂2 2. Maybe reduction must generate scalar output and thus the enclosing? Not 100% sure.

nice

then this should be correct.

@jimfan to remove the box, just used ⊃

@Razetime Yes and yours solution is 9-char only, even more compact

I've been thinking of using | and ⍣ for Euclid algorithm, no luck

Euclid algorithm?

lemme search that up

looks like a recursive func

@jimfan probably longer

@jimfan hint: ⌽|\

Okay let me try during daytime

@Razetime That's what I think too, the prime factor intersection is quite elegant

@ngn didn't expect ⌽ to be there

@Razetime if you do this repeatedly to a pair of ints, eventually you'll get to a pair of 0 and the gcd

the interesting part is figuring out how to avoid an infinite loop that toggles between 0,gcd and gcd,0

An interesting little sub-problem came up in "sporeball numbers," which seemed trivial but whose solution in J turned out to be surprisingly difficult. The problem is to make a verb that tells you if a list is a "doubled list" -- that is, some list repeated exactly twice. Any odd numbered list therefore always fails, and the empty list is supposed to fail too. The best I came up with was (*@#*]-:-:@#|.]) ::0. Can this be improved?

Try it online!

@Jonah I think -:]&.:(,~^:_1) works, although it returns 1 for an empty list while your solution doesn't. But why wouldn't an empty list qualify? It's the same as itself repeated twice.

<moon-child> @Marshall I expect because that's the requirement put forth in the ppcg question - codegolf.stackexchange.com/questions/211292

> The empty string "" is not palindromic, and cannot be split into two identical halves.

Well that's just not true...